What's Changed About AWPing In CS2?
Looking at some of the data from the conclusion of CS2's first top level international LAN at Sydney and offering some early answers as to whether the AWP really is "worse", plus exposing AWP abusers.
Preamble
IEM Sydney has just concluded, and FaZe Clan have taken the trophy in a stunning run. I’ll leave my general event conclusions for another article, seeing as it was my first LAN event and I do have a lot to talk about beyond the scope of this current topic.
One of the common talking points from pros and plebs alike is the AWP being different in CS2. Some say it is worse - others say it requires a change in playstyle - yet others will claim it is simply bugged somehow, though I think that last option is furthest from reality. The fact is that a number of intangible and minor changes to how the game feels in CS2 have changed the gun’s feel and style.
IEM Sydney saw the usage rate of the AWP drop to a level it has not been at since the earliest stages of CS:GO, but it also saw players like m0NESY and broky still finding form - so I’ve decided to break down some of these stats. I’ll be ignoring the four last place teams due to sample size (yes, this means I won’t be trash talking ZywOo) and seeing as NaVi didn’t really have an AWPer, that means a total of eleven AWPers will be under the microscope here.
If you want to keep up to date with my random ramblings on CS esports and other stuff, feel free to follow my Twitter!
General Conclusions
Before I inundate you with scatter plots, let’s discuss some of the more obvious overarching pieces of information.
AWPing is, on average, becoming less common. I wouldn’t have a way to check how many AWPs were purchased, but amongst the 11 AWPers tested today, the average has dropped from 51% to 35%.
This is actually fairly reasonable if you crunch the numbers. The number of rounds in a half has gone from 15 to 12 - if we scale down the AWP usage statistic we would get around 41%, which is somewhere in the middle. Nothing conclusive, but it helps paint the picture.
There’s also a lot of other stats which people might be clamouring for - the smoking guns that prove how bad the AWP in CS2 is. Sadly, I’ve got quite the opposite for averages, starting with some openings stats.
Maybe there’s truth to peeker’s advantage in general, but you won’t find it here - there’s really no change in how often these AWPers win their opening duels, nor how frequently they attempt to take them. Similar statements can be made about the ever-lucrative HLTV Rating.
Nothing significant enough to warrant further broad analysis, in my opinion. Where it gets interesting is when you look at how each player has changed their playstyle - so let’s move on to those specifics.
Reliance on the AWP for their precious HLTV Rating?
Let’s start with the obvious before we get speculative; m0NESY is the best CS2 AWP we’ve got. Despite G2 floundering against Complexity, he showed strong form throughout the event in both rifling and AWPing. The same could not be said for sh1ro, and I’ve got a few guesses as to why that you’ll see soon. Similar praise should be given to broky (both him and m0NESY are good riflers already, so it makes sense!), and similar criticism onto zorte, who only used an AWP for 22% of his kills.
Time to dump a scatter plot onto you!
Apart from afro, who remains an anomaly that I’m keeping my eye on in CS2, everyone reduced the proportion of kills they produced with the sniper rifle. This graph is effectively showing you who is performing despite that, and who has gotten worse as a result of it.
This starts to paint the picture. Dare I say it - sh1ro may be an AWP abuser. There were certainly other factors contributing to his performance being worse, like the change in leadership, but this is his first event below a 1.00 HLTV Rating since DreamHack Sevilla 2019. I’d be surprised if the inability to rely on an AWP as often is wholly irrelevant in this statistical cocktail.
Similar words, albeit praise instead of criticism, can be pointed towards m0NESY and broky. Most people knew that they were capable riflers, but it seems this will be a very important skill for AWPers if they want to stay on teams in the future.
Is aggression the key to performance for CS2 AWPers?
The short answer is maybe; the long answer is it depends on your skill set and it is hard to give a statistical answer with a limited sample size.
About half of the AWPers in CS2 actually took less opening duels than before (this is combining CT and T data). If I had to guess, I would say that with less AWPs to take specific duels with, they opted to let the regular riflers in their team (for instance, jkaem for nawwk, Keoz for acoR) take the duels more often, as they are more comfortable there.
More interesting than that is the colourful modifications to the graph I’ve made to illustrate some points. sh1ro and zorte, the two AWPers who experienced the largest drop off (and perhaps the two AWP abusers), have been taking more opening duels than they usually do, and are suffering for it. In particular, sh1ro is known for his passive playstyle - for now, that style of play may simply not work effectively.
The two within the light green, afro and hallzerk, are both fairly similar from the eye test, often happy to combat AWP. They’ve upped the aggression a lot, but are just about maintaining their existing form. I would take this as a tentative measuring stick for aggressiveness - previously, you would categorise these two as aggressive AWPers. Now, they’ve had to up the ante just to keep themselves the same, suggesting this is the new model for an aggro CS2 AWP.
The dot at the end, m0NESY, is an impressive anomaly. He already took a lot of opening duels, but has increased that even more (from 21.5% to 24.3%) in CS2, and it’s working for him. Naturally, his intuition and hundreds of hours are the primary assisting tools here, but I think it’s no coincidence that he’s being more aggressive than before.
Conclusions?
This sample size is quite small, but some early understandings can be reached.
This is Pearson’s correlation coefficient, comparing the correlation between opening kill attempts and HLTV Rating prior to Sydney and during Sydney for these AWP players. While it’s a weak correlation, there is a marked change - a small negative correlation has turned into a small positive correlation. More data is needed to be conclusively sure on this, but my eye test tells me the same thing - aggressive AWP plays are the current meta of CS2.
Here’s where I give you the disclaimer - this was a small amount of maps at the first ever big CS2 LAN and the compounding factors that affect something like HLTV Rating are off the charts. sh1ro could literally just have had a few bad days of sleep from his flight and made some bad plays that resulted in a skewed sample size for people to criticise him with. The existence of MR12 and the subtick system changing how you are supposed to aim is not easy to measure numerically at all.
Nevertheless, I think the basic conclusion is that passive AWP players need to adapt their style, or at the very least become adept riflers in their own right, and aggressive AWP players are salivating with CS2’s mechanics offering them better opportunities.
Afterword
Thank you for reading this article all the way to the end! I’ve got another one in the works, talking about the good and bad of IEM Sydney (mostly good!), so stay tuned for that, and shoot me a Tweet or DM over there if you want to talk CS.
If you’re in the mood for some Super Auto Pets gameplay, I’ve been uploading pretty much daily on my YouTube channel - it’s not CS, but it’s fun nonetheless, so check it out.